Thorax 1982;37:175-180

Effects on pulmonary function of whole lungirradiation for Wilm's tumour in children

MR BENOIST, J LEMERLE, R JEAN, P RUFIN, P SCHEINMANN, J PAUPE

From the Laboratoire d'Exploration F'onctionnelle Respiratoire, H6pital des Enfants Malades, Service du PrJ Paupe, and Institut Gustave Roussy, Service de Pediatrie, Villejuif, France

ABSTRACT The effect of whole lung irradiation on lung function was investigated in 48 childrentreated for Wilm's tumour with pulmonary metastases. Lung function tests were performed beforeirradiation and were repeated annually for as long as possible, the length of follow-up varying fromtwo to 17 years. A reduction in both lung volume and in dynamic compliance was clearly observed.In some patients these changes occurred in the early post-irradiation months, but in most thedecrease observed progressed over longer periods of time. Static pressure volume curves, blood-gases, and carbon monoxide transfer were normal. These findings make it unlikely that post-irradiation pulmonary fibrosis was involved. Another explanation for the decreased lung volumeand dynamic compliance might be failure of alveolar multiplication. Muscular injury is unlikely asthe patients were able to produce normal transthoracic pressures. A failure of chest wall growth isalso possible and would explain the progressive restrictive impairment but not the early lungfunction changes. It is suggested that the early effects detected in some patients were the result oflung injury and that later effects resulted from impaired chest wall growth.

Although the effects of pulmonary irradiation onlung function have been extensively investigated inadults, reports on the effects on respiratory functionof whole lung irradiation are relatively few and therehave been no reports of detailed, repeated lungfunction tests after this type of irradiation.

In contrast to adults, young children are in aperiod of rapid lung and skeletal growth and theeffect of pulmonary irradiation might be a failureof alveolar development resulting from impairedcellular proliferation, thus decreasing the number ofalveoli. The aim of this study was to detect the short-and long-term pulmonary function changes in youngchildren who received whole lung irradiation at doseswhich were therapeutically efficient but close totolerance and we have tried to characterise thephysiopathological mechanisms of the changesobserved. Actinomycin D which enhances the effectsof irradiation, was administered consecutively.1

Address for reprint requests: Dr MR Benoist, Laboratoired'Exploration Fonctionnelle Respiratoire, Service du Pr JPaupe, H6pital des Enfants Malades, 149 rue de Sevres,75730 Paris Cedex 15, France.

Methods

Forty-eight children undergoing treatment forWilm's tumour associated with pulmonary met-astases at the Institut Gustave Roussy between 1960and 1976 were studied. Clinical and therapeuticcharacteristics are summarised in table 1. In 39 ofthese patients, nephrectomy was followed by ir-radiation of the tumour bed, including the entirewidth of the spine, the field extending to the hemi-diaphragm. All the children received bilateralpulmonary irradiation using 60Co at a dose of 20Grays/three weeks. Actinomycin D was always givenin conjunction.Twenty-two patients were studied prospectively

with initial lung function testing carried out justbefore pulmonary irradiation and this was denotedtest "O". This information was lacking for the other26 children who were included in the study whichbegan in 1973. All 48 patients were subsequentlytested in the same way. Measurements of lungfunction were repeated in each patient every sixmonths during the first two years after pulmonaryirradiation and then annually for as long as possible,

175

Benoist, Lemerle, Jean, Rufin, Scheinmann, Paupe

Table 1 Clinical information

Number of patients 48 (25 M, 23 F)Mean age at diagnosis 3 years ± 9 months (ranged from

I to 12 years)Irradiation to tumour bed.

Midplane exposure 30 to 50 Grays*Bilateral pulmonary irradiation.

Midplane exposure 20 Grays/21 daysMean age at pulmonary 4 years ± 7 months (ranged from

irradiation I to 13 years)Metastases Bilateral in 38 patients, unilateral

in 10 patients (disappeared inall cases before test "O")

Chemotherapy 48 patients: Actinomycin D(15 Mg/kg. 5 during tumour bedand pulmonary irradiation)

Long-term follow-up 7 years ± 3 months (2 to 17years)

*J Gray = 10-2rads.

At present the length of the follow-up varies fromtwo to 17 years.When the cooperation of the child permitted, total

lung capacity (TLC) and vital capacity (VC) weremeasured using a nine-litre Godart respirometer.Functional residual capacity (FRC) was determinedby duplicate measurements using the helium dilutionmethod. Volumes (BTPS) were expressed as a per-centage of predicted normal values.2Dynamic lung compliance (CLdyn) and pulmonary

resistance (RL) were determined by means of theoesophageal balloon technique.3 Length, circum-ference, and position were all specified in a previousstudy,4 the difference between mouth and oeso-phageal pressure being measured with a Fenyves-Gutdifferential pressure transducer, and flow wasmeasured with a Fleish pneumotachograph (number1, 2, or 3) coupled to a Fenyves-Gut pressuretransducer. Gas volumes were determined byelectrical integration of the flow signal. Flow,volume, and oesophageal pressure changes wererecorded on a three channel recorder (Rapidgraph,Sefram RPS) with a linear response curve up to 10Hz. CLdyn was calculated from the ratio of tidalvolume to the change in transpulmonary pressurebetween the points of zero air flow between eachinspiration and expiration. Each CLdyn value wasaveraged from at least 10 successive breaths atresting tidal volume. RL was determined by themethod of Mead et al,5 and these values were alsoexpressed as a percentage of the predicted normalvalues according to height.Lung elasticity was assessed in 18 subjects using

measurements of static lung compliance (CLst) andelastic recoil pressure of the lungs (Pst). Staticdeflation pressure-volume (PV) curves were obtainedby interrupting the expiratory airflow from TLC toresidual volume (RV) for five to six four-secondperiods. Constant volume conditions were achievedby preceding each recorded curve with three full

inflations. Transpulmonary pressure was plottedagainst lung volume expressed as percentage ofmeasured TLC and the slope of the PV curve in thetidal volume range gave the static expiratory lungcompliance (CLEst). The values of Pst at variouspercentages of TLC (eg Pst insp relaxed 100, 90, 80,and so on) was obtained from the PV curves.6 7 Aline of best fit was hand drawn through at least threesets of PV data which differed by only ± 1 cm H20.Data were compared to those obtained in 23 controlsubjects of the same age.2Blood oxygen tension (Pao2) was measured using

arterialised capillary blood collected from the fingertip and was compared with predicted values similarlyobtained.2 It was possible to measure Pao2 beforeand after exercise in 14 co-operative subjects usinga bicycle ergometer (Monark). The exercise wasperformed for six minutes at successive loads of one-

.... -. .....

_ ~~~~~~~~~~~~~~..M E~~~~~~~~~~~~~~~~~~~.~~~~~~~~~~~~~~...

.~~~~~~~~~~~~~~X.S. .... .

Fig 1 Thoracic hypoplasia seen in a 9-year-old girl.

176

Effects on pulmonary function of whoke lung irradiation for Wilm's tumour in children

Table 2 Lung volumes, pulmonary mechanics (mean of expected values and SD) and Pao2 (arterialised capillaryblood) in 1I patients followedfrom admission to the fourth year after irradation

Test FRC VC TLC CLdyn RL Pao2 mmHg

M SD M SD M SD M SD M SD M SD

"0" 97 14 - - 90 19 103 20 88 56 months 95 15 - - 68t 14 98 18 85 412 months 91 16 - - 70t 17 99 11 84 418 months 89* 10 78 12 80 13 74 14 97 13 86 42 years 90* 17 78 8 79 6 73* 13 91 9 92 63 years 78t 12 72 6 74 9 60t 8 90 12 90 54 years 71t 9 70* 9 72* 7 60t 9 85 8 88 7

*p < 0 05; tP < 0-02 (paired I test "O" versus successive tests).

third and two-thirds of the maximum predicted forthe height of the child.8

Transfer factor for carbon monoxide (TLCO) was

measured in 13 subjects by the single breath method.9

Results

CLINICALSymptoms of radiation pneumonitis developedacutely two to three weeks after completion ofradiation therapy in two of the 48 children and was

characterised by non-productive cough, dyspnoea,fever and in one case progression to severe respir-atory distress. The symptoms and radiographicfeatures subsided after two months for the latterpatient, and after three weeks for the remainder. Onclinical examination greatly reduced sagittal andfrontal thoracic diameters were observed in nearlyall of the cases seen three or four years after radiationtherapy (fig 1). Shortening and deformity of thelower thoracic and lumbar vertebrae within theabdominal treatment field were also observed afterthis follow-up for three or four years. Spinalcurvatures ranging from 10 to 30 degrees were

observed in 10 of the 48 subjects but these changesdid not influence height, as only two subjects were

below two standard deviations and one was even

above the mean height for his age. The patients were

not disabled and most of them experienced no

dyspnoea on moderate exertion.The chest radiographs showed no pleural or

pulmonary abnormalities after the first two yearsafter pulmonary irradiation.

MEASURED LUNG FUNCTION

As it was only possible to study some of the 48patients over a long period, the data obtained are

presented in two different ways.

First, table 2 summarises the results in 11 subjectswho were followed for four years. Lung volumes andCLdyn, expressed as the mean of the percentage ofthe predicted values, decreased with time. Functionalresidual capacity decreased from 97% to 71 %(p < 002), VC from 78% to 70% (p < 0-05) andCLdyn from 90% to 60% (p < 0'02). The reductionwas more significant at the three and four year tests(p < 0'02) than at the one year test (p < 0'05).

100-

90-

80-

E u 70-do

cL 5 60-

-y 50-

40-

?20-

0.."°.......

S.,.....

TLC,:I

/Ldyn

0 1 2 3 4Successive testng (tme n years)

5 6

Fig 2 Percentage ofpatients with significantly reducedTLC and CLdvn plotted against successive tests in all 48subjects.

Table 3 Lung volumes, pulmonary mechanics (mean of expected values and SD) and Pao2 (mmHg, arterialisedcapillary blood) obtained over test "O" to > six years period in all 48 subjects (part 2).

Tests Number FRC VC TLC CLdyn RL Pao,ofsubjects M SD M SD M SD M SD M SD M SD

"0" 22 105 4 - - 86 31 102 19 89 66 months 18 91 11 - - 72 29 105 16 87 718 months 17 92 12 76 11 76 9 72* 11 99 15 89 64 years 18 78t 14 70t 16 70t 13 62t 10 92 9 90 5

a 6 years 15 75t 11 61t 13 66t 13 50t 13 85 7 91 8

Full data may be obtained from the authors.*p < 0 05; tP < 0-02. Analysis of variance "O" versus successive tests.

I I I I

177

178

o80- fD

E

- 70-

E260-

50-C,,

400 10 20 30 40

Transpulmonory Pressure (cm H20)

Fig 3 Mean pressure-volume curves in 18 patients, lungvolume expressed as a percentage of actual TLC

±-)± I SEM and in 23 controls (-------

10 -Control

60

-J~~~~~~~~~~~~~~-----

0 10 20 30 40Trarspidmonary Pressure (cm H20)

Fig 4 Pressure-volunme curves in four patients. Volunmeexpressed as per-centage of actual TLC (as percentage ofpredicted TLC (-------

Secondly, the complete data obtained from all 48subjects by repeated testing are shown in table 3. Theperiod of follow-up varied between patients eitherbecause of mortality or because of recruitment forthe study at a later date. In this group too, we

observed a similar significant reduction in the mean

values with the passage of time after pulmonaryirradiation. Functional residual capacity decreasedfrom 105% to 75% (p < 002), TLC from 76o% to66% (p < 002), VC from 76% to 61% (p < 002),

Bentoist, Lemerle, Jeani, Rufilt, Scheinmaniii, Paupe

Table 4 Results of mechanical factor-s in 18 subjectstested at five years ± three months after lung irrladiationcompared with 23 controls

CLdyn Pst (ciii H20) CLESIx 100

CLESt 100 90 80 70 60 VC% oJ TLC m1ieasured

PatientsMean 81 34 26 17 12 8 0 033SD 16 8 9 5 2 8 2 0-005ControlsMean 85 37 25 16 10 8 0 038SD 13 6 5 3 2 5 3 0-007

I Cm H20 = 0 1 kPa.

and CLdyn from 86% to 50% (p < 002). Thenumber of patients with reduced values increasedwith time after irradiation (fig 2).

In the tests taken as a whole, there were morereduced values of VC and TLC than FRC(p < 0 001) and the decrease in CLdyn was morepronounced than that of lung volume (p < 0 001).The RL values were not significantly different fromthe predicted values for every patient at each testing(tables 2, 3). Pressure-volume curves obtained for 18patients after five ± 2 9 years follow-up were nodifferent from those of 23 controls when Pst was

Table 5 Results of lung transfer .fr CO and PaQ2before and after exercise

TLCO (SB) % TLCO (SB)', Pao,. torrsof expected of expectedfor height corrected for Before After

lung volunie exercise exercise

Mean 72 112 85 87SD 8 11 6 7n 13 13 14 14Years elapsed 6± 4 6± 4 6± 4 6 ± 4

TLCO (SB) = transfer capacity for CO, single breath method;n = number of subjects; kPa = 7 5 torrs.

related to measured lung volume (fig 3) but whencompared with expected lung volume, the curveswere found to be displaced downwards and to theright (fig 4).Transpulmonary pressure measured at various

percentages of TLC was not increased (table 4) andCLEst was higher than CLdyn, as is usual in healthychildren.'0 It The mean ratio of CLEst with VC wasnot significantly lower than in healthy children(table 4).

Arterial oxygen tension was within the normallimits on every test (tables 2, 3). Post-exercise Pao2was not reduced in 14 children tested after six years(table 5).Mean TLCO was found to be significantly reduced

in 13 patients tested after six years follow-up:72 ± 8% of the predicted value for height(p < 0 02), but was no lower than predicted for lungvolume: 112 ± 11 % (table 5).

I I~~~~~~~~~~~~

Effects on pulmonary function of whole lung irradiation for Wilm's tumour in children

_ 90-

'a

; 70-C

50-

....... -.0. 0,,,,,,,.

~~~ ~ ~ ~ ~ ~ ~ ~ .1

Ap

0 6moI ~ ~ I

18mosuccessive tests

Fig 5 Time-course of changes in CLdyn in I(0), in nine subjects with early fall in CLdyand in nine subjects with initially preserved v

Discussion

A restrictive ventilatory impairment ahypoplasia were clearly observed in theour study. This decrease in lung volumeto that found by Wohl12 in six patientsto 13 years after undergoing similar trwith lower radiation exposure.

Three chronological phases could be dpresent study. The first, observed duripost-irradiation months, was charadeteriorating lung function. The secondnext two years when no further deteriseen. The third phase occurred after theand was characterised by both progicreasing mean values and an increasecpatients with abnormal lung volume an

2).These chronological changes may be

different pathological processes, the firsearly lung disease. Thus at the six-monnine of the 18 patients tested showed a d(mean 48% of predicted, range 30% tcof them relapsed with severe pulmonaxand died some months later withdistress symptoms, while two othincreased bronchovascular markingsblurring on the chest radiograph and

had evidence of interstitial pneumonitis. Howeverno explanation for the deterioration in lung functioncould be found in the other three subjects. In theremaining nine patients tested at six months, no

.0 deterioration in CLdyn was observed (mean CLdyn95% of predicted, range 111 to 72%).One year later, at the 18-month test (fig 5), the

mean CLdyn of the seven surviving subjects hadimproved (mean CLdyn 67% of predicted). It islikely that the apparent stability of the mean valuesobserved during the second phase was caused by acombination of improved lung function in somesubjects, and omission of low values as a result of

I i""'r deaths in two others. During the third phase,y progressive deterioration could be explained by some

allpatients lung parenchymal sequelae of acute lung injury andl p ) by the deleterious effects of radiation on lung and/or

values (O). chest-wall growth.~ Several physiopathological mechanisms might

explain these persistent changes. Pulmonary fibrosiswas considered as a possibility; however as pre-viously seen only two subjects showed evidence of

ind thoracic interstitial pneumonitis and we found no increase ine children in Pst (L) measured after five years in 18 subjects, ofwas similar whom five exhibited early lung function deteriorationtested seven such as is found in adults with pulmonary fibrosis.eatment but Elastic pressures have been measured by Yernault,13

Macklemi,'4 West,'5 and Gibson'6 at differentlefined in the measured lung volumes and the static PV curvesing the early were clearly shifted downwards and to the right whencterised by compared with normal curves. In contrast, the PVspanned the curves in our 18 subjects remained similar to thoseioration was of healthy subjects. Two other findings make fibrosissecond year unlikely: first, post-exercise Pao2 was no lower than

ressively de- that before exercise in 14 patients, three of whomd number of exhibited greatly decreased CLdyn at the six-monthId CLdyn (fig test; secondly, TLCO was normal in all measurements

when related to the measured TLC.the result of Another possible explanation for decreasedt of which is volume and CLdyn might lie in injury to alveoli andth test (fig 5) peripheral airways which could have been destroyedrop in CLdyn or failed to develop at this age when there is normallyv 58 %). Two intensive alveolar multiplication. The more markedry metastases reduction in CLdyn compared with lung volume may

respiratory be caused by ventilatory asynchrony resulting fromiers showed injury, with unequal changes in the calibre of theand basal peripheral airways. To the extent that reduction in

two patients size or number of alveoli may be comparable to a

Table 6 Lung volume and CLdyn in 17 children with renal treatment but without pulmonary irradiation

Tests in years CLdyn FRC VC TLC

Mean SD Mean % SD Mean % SD Mean % SD Mean % SDof of of ofexpected expected expected expected

5 4 92 8 96 12 83 7 88 91

I

179

Benoist, Lemerle, Jean, Rufin, Scheinmanai, Paupe

pneumonectomy, CLdyn might be reduced accordingto reduction in lung volume. It may be seen from thePV curves that, at normal predicted pressure, lungvolume was smaller than expected. This findingsuggests a reduction in the number of alveoli. Analteration in lung growth may induce a parallelalteration in chest-wall growth, leading to a smalllung in a small thorax.

Alternatively, it is possible that impairment insome subjects was caused primarily by muscular orskeletal injury. Muscular damage to the diaphragmfrom radiotherapy might occur after the irradiationof the tumour bed but in 17 subjects with similarrenal treatment without pulmonary irradiation, nosignificantly reduced values were observed (table 6).Intercostal muscle weakness is also a possibleconsequence of thoracic irradiation. We did notrecord the maximum static respiratory pressures,recently described by De Troyer17 and requiringgood motivation and cooperation from the subjects,but it is unlikely that respiratory muscle strength wasdecreased in subjects who were able to performnormal Pst. Skeletal radiation damage identical tothe vertebral body changes seen after the kidneytumour bed irradiation is also possible.18 19 Ifthoracic hypoplasia, which increases with the passageof time, is caused by failure of chest-wall growth,this could explain the progressive diminution of lungvolume but not the early functional changes.

It is therefore likely that pulmonary irradiationhad direct consequences, either on the pulmonaryparenchyma or on skeletal growth and that in someof the patients tested, these various mechanisms maybe involved in functional pulmonary abnormalities.In some patients the early primary phenomenon wasthat of lung injury, with deleterious later effects onboth lung and skeletal structures, while in otherpatients a reduction in chest-wall growth might besolely responsible for these changes.Our results in these patients suggest relatively well-

preserved function despite a whole lung radiationdose which was close to the limit of tolerance.Nevertheless, it must be emphasised that spinalcurvature may deteriorate at puberty, causing furtherimpairment in lung function, so that continuedfollow-up up to and beyond puberty is essential.

References

Margolis LW, Phillips TL. Whole lung irradiation formetastatic tumor. Radiology 1969 ;93 :1173-9.

2 Jean R, Benoist MR. Exploration fonctionnelle en pediatrie.First edition. Paris: Flammarion, 1979.

3 Milic-Emili G, Mead J, Turner JM, Glauser EM. Improvedtechnique for estimating pleural pressure from oeso-phageal balloons. J Appl Physiol 1964;19:207-1 1.

4 Benoist MR, Siguier C, Fermanian J, Jean R, Paupe J,Vialatte J. Lung function after neonatal respiratorydistress syndrome. Bull Europ Physiopathol Respir 1976;12:703-14.

' Mead J, Whittenberger JL. Clinical measurement ofresistive and elastic properties of the lung. J Clin Invest1953 ;32 :588-9.

6 Colebatch HJH, Greaves IA,Ng EKY. Exponential analysisof elastic recoil and aging in healthy males and females.J Appl Physiol' 1979;47:683-91.

Finucane KE, Colebatch HJH. Elastic behaviour of thelung in patients with airway obstruction. J Appl Physiol1969 ;26:330-43.

Godfrey S. Exercise testing in children. Philadelphia:Saunders, 1974.

Ogilvie CM, Forster RE, Blakemore WS, Morton JW. Astandardised breath holding technique for the clinicalmeasurement of the diffusing capacity of the lung forcarbon monoxide. J Clin Invest 1957;36:1-17.

10 Van der Hardt H, Logvinoff MM, Dickreiter S, GeubelleF. Static recoil of the lungs and static compliance inhealthy children. Respiration 1975;32:325-39.

Zapletal A, Paul T, Samanek M. Pulmonary elasticity inchildren and adolescents. J Appl Physiol 1976;40:953-61.

12 Wohl ME, Griscom NT, Traggis DG, Joffe N. Effects oftherapeutic irradiation delivered in early childhood uponsubsequent lung function. Pediatrics 1975;55:507-16.

13 Yernault JC, De Jonghe M, De Coster A, Englert M.Pulmonary mechanics in diffuse fibrosing alveolitis. BullEurop PhYsiopathol Respir 1975;11 :231-44.

14Macklem PT, Becklake MR. The relationship between themechanical and diffusing properties of the lung in healthand disease. Am Rev Respir Dis 1963;87:47-58.

15 West JR, Alexander JK. Studies on respiratory mechanicsand the work of breathing in pulmonary fibrosis. Am JMed 1959 ;27:529-44.

16 Gibson GJ, Pride NB. Lung distensibility in pulmonaryfibrosis. Bull Europ Physiopathol Respir 1973;9:1250-1.

17 De Troyer A, Yernault JC. Inspiratory muscle force innormal subjects and patients with interstitial lungdisease. Thorax 1980;35:92-100.

18 Oliver JH, Gluck G, Gledhill RB, Chevalier L. Musculo-skeletal deformities following treatment of Wilm'stumor. Can Med Assoc J 1978 ;119:459-64.

19 Heaston DK, Libshitz HI, Chan RC. Skeletal effects ofmegavoltage irradiation in survivors of Wilm's tumor.AJR 1979;133:389-95.

180